Modeling and parameter identification of microbial batch fermentation under environmental disturbances

Abstract

This paper considers the mathematical modeling and parameter identification of glycerol bioconversion to 1,3-Propanediol in the batch fermentation. Taking into account the characteristic of inherently genetic stochasticity and environmental disturbances in this biochemical system, a stochastic differential system with disturbance is proposed to describe this microbial fermentation process. To guarantee the model refinement and the biological robustness simultaneously, a multi-objective parameter identification model is proposed. To obtain the optimal parameters, the multi-objective optimization problem is converted to a sequence of single-objective optimization problems. Then, the problems are further approximated to a sequence of the optimization problems with linear matrix inequality constraints under the help of the T-S fuzzy approximation. Finally, the optimization problems were solved by using the linear matrix inequality optimization techniques. Numerical results show that the proposed stochastic dynamical system is fit to describe this batch process and our proposed method of parameter identification is also feasible.